US7987150B1 - Method and apparatus for automated rule-based sourcing of substrate microfabrication defects - Google Patents
Method and apparatus for automated rule-based sourcing of substrate microfabrication defects Download PDFInfo
- Publication number
- US7987150B1 US7987150B1 US11/849,869 US84986907A US7987150B1 US 7987150 B1 US7987150 B1 US 7987150B1 US 84986907 A US84986907 A US 84986907A US 7987150 B1 US7987150 B1 US 7987150B1
- Authority
- US
- United States
- Prior art keywords
- defect
- defects
- density
- cluster
- clusters
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Expired - Fee Related, expires
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N20/00—Machine learning
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N5/00—Computing arrangements using knowledge-based models
- G06N5/02—Knowledge representation; Symbolic representation
- G06N5/022—Knowledge engineering; Knowledge acquisition
- G06N5/025—Extracting rules from data
-
- H—ELECTRICITY
- H01—ELECTRIC ELEMENTS
- H01L—SEMICONDUCTOR DEVICES NOT COVERED BY CLASS H10
- H01L22/00—Testing or measuring during manufacture or treatment; Reliability measurements, i.e. testing of parts without further processing to modify the parts as such; Structural arrangements therefor
- H01L22/20—Sequence of activities consisting of a plurality of measurements, corrections, marking or sorting steps
-
- H—ELECTRICITY
- H01—ELECTRIC ELEMENTS
- H01L—SEMICONDUCTOR DEVICES NOT COVERED BY CLASS H10
- H01L22/00—Testing or measuring during manufacture or treatment; Reliability measurements, i.e. testing of parts without further processing to modify the parts as such; Structural arrangements therefor
- H01L22/10—Measuring as part of the manufacturing process
- H01L22/12—Measuring as part of the manufacturing process for structural parameters, e.g. thickness, line width, refractive index, temperature, warp, bond strength, defects, optical inspection, electrical measurement of structural dimensions, metallurgic measurement of diffusions
Definitions
- the present invention relates to detection of semiconductor fabrication defects, and in particular to a method and apparatus for automatic rule-based semiconductor fabrication defect signature recognition and defect sourcing.
- An automated defect sourcing system identifies the root-cause of process excursions due to contamination, process faults, equipment failure and/or handling.
- the system analyzes defect data in timely manner and provides accurate timely feedback to address and contain the sources of yield excursions.
- a signature repository stores descriptions of known wafer surface manufacturing defects as set of rules.
- the signature of a manufacturing defect pattern is associated with a type of equipment or process, and is used to source the manufacturing defects and to provide process control for changing and/or stopping yield excursions during fabrication.
- a defect signature rule-based engine matches the data in a wafer defect inspection file against the set of rules in the signature repository during wafer fabrication. Once the defect signature is detected during fabrication, handling and/or disposing the root-cause of the corresponding defect signature is facilitated using messages according to an event handling database.
- a real-time process control for wafer fabrication is provided.
- FIG. 1 is a flow diagram illustrating a rule-based method for defect sourcing, according to an embodiment of the present invention.
- FIG. 1 a is a flow diagram illustrating a method for rule extraction, according to an embodiment of the present invention.
- FIG. 1 b shows an example of defects on a wafer defect map, according to an embodiment of the present invention.
- FIG. 1 c shows an enlargement of a portion of a wafer defect map, showing defects that share spatial proximity, with a virtual grid superimposed on the defects, according to an embodiment of the present invention.
- FIG. 1 d shows active units within the superimposed virtual grid that may be grouped into a cluster, according to an embodiment of the present invention.
- FIG. 1 e shows a cluster of defects that share spatial proximity, according to an embodiment of the present invention.
- FIG. 2 a shows some basic features of an ellipse density model represented by a spine or major axis, a width or minor axis, and a centroid, according to an embodiment of the present invention.
- FIG. 2 b shows a graphical user interface with an ellipse density model superimposed on a defect cluster contour, according to an embodiment of the present invention.
- FIG. 2 c shows a curved tube density model superimposed on a defect cluster, according to an embodiment of the present invention.
- FIG. 2 d shows a rectangle density model superimposed on a defect cluster, according to an embodiment of the present invention.
- FIG. 2 e shows a polygon density model superimposed on a defect cluster, according to an embodiment of the present invention.
- FIG. 2 f shows several density model shapes for clustering defects, according to an embodiment of the present invention.
- FIG. 2 g shows convex and non-convex clusters, according to an embodiment of the present invention.
- FIG. 3 is a flow diagram illustrating a method for zonal analysis and the rule creation, according to an embodiment of the present invention.
- FIG. 3 a shows a graphical user interface for a zone editor tool used to define a zone in the wafer defect map analysis and classification, according to an embodiment of the present invention.
- FIG. 4 a shows a group of defect clusters sharing the same column topology, according to an embodiment of the present invention.
- FIG. 4 b shows a group of defect clusters sharing the same row topology, according to an embodiment of the present invention.
- FIG. 4 c shows a group of defect clusters sharing the same annular topology, according to an embodiment of the present invention.
- FIG. 4 d shows a group of defect clusters sharing the same radial topology, according to an embodiment of the present invention.
- FIG. 4 e shows a group of defect clusters sharing the same off-center linear topology, according to an embodiment of the present invention.
- FIG. 4 f shows a group of defect clusters sharing the same spiral topology, according to an embodiment of the present invention.
- Defect A pit, tear, groove, inclusion, grain boundary, or other surface feature anomaly that is either characteristic of the material or a result of its processing and that is not a result of the sample preparation. Examples include: particle deposited on substrate; scratch or streak on wafer; lithographic error; non-uniform deposition, etch or planarization; excessive or inadequate deposition, etch or planarization; etc.
- Defect Cluster A grouping of defects detected on a wafer surface that have commonality in position, size, orientation and/or one or more other measurable attributes.
- Scratch cluster A grouping of concentrated defects that comprises a scratch and spans multiple dies. When an envelope is drawn around the scratch cluster, the envelope shape is generally elongated; however, the position and orientation of the cluster is such that it is not radial or annular. There are two types of scratch clusters: random and patterned.
- Patterned scratch A classification of scratch defect clusters. Patterned scratches comprise multiple scratch clusters that are arranged in a recognizable pattern.
- Annular Defect Cluster A grouping of distributed defect clusters. Annular clusters are distributed in a ring pattern with approximately equal distance from the center of the wafer. An annular cluster may comprise two or more rings (e.g. a “bull's eye”).
- Radial Defect Cluster A grouping of distributed defects. Radial clusters are distributed along lines emanating from the center of the wafer. A radial signature may also include a concentration of defects in the center of the wafer. Starfish clusters are examples of radial clusters.
- Zonal Defect Cluster A grouping of distributed defects that span multiple dies. Zonal clusters are the least well-defined; they may be characterized by location, shape, orientation, density distribution and relationship to other clusters. There are two classifications of zonal defect clusters: fixed position and floating.
- Inspection Equipment Equipment used by a manufacturer to identify defects, artifacts and/or anomalies on a substrate surface and/or below a substrate surface. Inspection equipment may be stand-alone or integrated into the process equipment (i.e. in situ).
- Metrology Equipment Equipment used by manufacturers to measure topography, morphology, film thickness and/or other physical attributes of the substrate surface and/or attributes that appear below the substrate surface. Metrology equipment may be stand-alone or integrated into the process equipment (i.e. in situ).
- Results File The output file (such as a text file, digital image file or wafer defect map) from the inspection equipment or the metrology equipment.
- Inspection equipment results files define or indicate the coordinate location, size, type and/or other attributes of defects, artifacts or anomalies. Examples include KLATM results files, Scanning Electron Microscope images, wafer bin maps, digital images of wafer surface, etc.
- a metrology equipment results file indicates physical attributes at various points over the wafer surface. Examples are film thickness uniformity maps, wafer reflectivity maps, etc.
- Wafer Defect Map or Substrate Defect Map A display of defect information corresponding to defect coordinate locations on the wafer or substrate surface.
- Defect Signature A pattern or collection of patterns that are consistently created by a degradation, excursion or malfunction in a specific manufacturing process step.
- Die stack Wafer inspection data from multiple dies and/or multiple wafer files overlaid onto a single die grid. The purpose of creating a die stack is to identify systematic problem areas within the die.
- Reticle stack Wafer inspection data from multiple reticles and/or multiple reticle files overlaid onto a single reticle grid. The purpose of creating a reticle stack is to identify systematic problem areas within the reticle.
- Hot spot A region of the wafer or die, which contains a disproportionately high percentage of defects.
- FIG. 1 is a flow diagram illustrating a rule-based method for defect signature sourcing, as well as its inputs and outputs.
- One or more result files 100 are used in process 200 to define or extract features of spatial defect clusters. The features are then used to manually or automatically build a set of rules 300 . These rules are used in process 400 to classify spatial defect signatures.
- One or more result files 100 are used in process 250 to manually define or automatically derive a set of rules based on a zonal analysis of the surface of the wafer defect map.
- Process 400 will read in one or more inspection result files during wafer fabrication processing steps, classify the spatial signatures according to the set of rules in library 300 , and output one or more classified defect result files 120 .
- process step 500 is used to optimize the spatial signature recognition rules in library 300 .
- FIG. 1 a is a flow diagram illustrating a method for rule extraction and generation.
- the rule is then stored in the object library.
- the rule extraction process 200 is described in terms of a graphical user interface (GUI), operating within a graphical computer operating system.
- GUI graphical user interface
- FIG. 1 b shows an example of defects 6 on a wafer defect map 5 .
- a GUI displays a wafer defect map 5 showing defects 6 to a defect engineer or other user.
- the user perceives a defect cluster that can serve as a signature.
- the user identifies a defect cluster through a trial and error process of adjusting density and distance parameters, as described below, until the cluster defined by the parameters sufficiently approximates the perceived defect cluster that is to serve as a signature. Once the density and distance parameters are adjusted to identify a defect cluster, a mathematical density shape model is produced to approximate the perceived defect cluster that is to serve as a signature.
- FIG. 1 c shows an enlargement of a wafer defect map, showing defects that share spatial proximity, with a virtual grid 10 superimposed on the defects 11 .
- the user adjusts density and distance parameters which result in grouping active cells (such as cells 12 , 14 and 15 , hereinafter also called units) into a cluster based on spatial proximity.
- the wafer defect map is divided into a virtual grid 10 where the grid cells may be measured in microns or nanometers depending on the application or resolution of the spatial signature.
- An active cell is defined as one having at least one defect 11 within the virtual grid cell.
- An empty cell 13 is defined as having no defect in the virtual grid cell.
- a cluster is formed when active cells share spatial proximity in cell neighbor distance (a cell is considered to have a cell neighbor distance of one to each of its eight surrounding neighbor cells).
- FIG. 1 d shows that active cells 22 , 23 , 24 , 25 , 27 , 28 and 29 within the superimposed virtual grid are grouped into a cluster of cells that have a cell neighbor distance of one. Active cells 21 , 26 and 30 are not part of the cluster because the cell neighbor distance to the nearest active cell is greater than one and the rule for the cluster is defined to be equal to one.
- FIG. 1 e shows an example cluster 32 with a neighbor distance of one, with individual defects 31 , 33 , 34 and 35 not considered to be part to the cluster 32 .
- a user specifies distance and density parameters and views a wafer map together with the clusters, defined by the parameters, superimposed thereon. Such clusters are hereinafter also referred to as “primitive clusters.” The user then goes through an iterative trial-and-error process of adjusting the density and distance parameters of the primitive clusters until they best agree with the defect cluster signature that the user desires to capture.
- a GUI allows the user to draw a boundary around the perceived defect cluster that is to serve as a signature, and the computer searches for distance and density parameter values that approximate the defect cluster.
- the density and distribution of defects within the primitive clusters are used to formulate one or more mathematical density shape models for the primitive clusters.
- Such density shape models represent approximations of the density characteristics of the primitive clusters.
- An advantage of working with such mathematical approximations is that they comprise a set of parameters (described below) which form a more concise basis for defining rules that describe defect clusters.
- a density shape model may be a best-fit ellipse which has the same second moment as an individual cluster, as shown by the following equation:
- the density shape model is an ellipse.
- FIG. 2 a shows basic features of an ellipse density shape model represented by a spine 56 or major axis, a width 57 or minor axis, and a centroid 55 .
- FIG. 2 b shows a graphical user interface with an ellipse density shape model 58 superimposed on the defect cluster contour 59 .
- the density shape model is not limited to an ellipse density shape model, and other shapes are possible as would be apparent to one skilled in the art.
- FIGS. 2 a , 2 c , 2 d and 2 e are models that share the following features:
- FIG. 2 f shows examples of feature shapes where 81 and 82 have straight spines, 85 , 87 and 89 have curved spines, 80 , 83 , 85 and 86 have cross-section symmetrical reflection, and 88 has asymmetrical reflection.
- step 205 extracts the spatial features of the individual density shapes in order to describe the defect cluster, such as described in the following:
- the rules can be generated to combine clusters based on parameters such as the clusters' linearity, proximity, density, radial position, horizontal or vertical position, etc.
- a series of three ellipses in a particular triangle formation may form a complex cluster.
- a series of four ellipses in a line formation may form a complex cluster.
- any combination of density shape models can be used to form a complex cluster.
- this iteration is not limited to two levels, but can be repeated as desired.
- FIG. 4 a shows examples of defect clusters in which the centroids of the density shape model, or centers of density mass, align along an x-axis in a Cartesian coordinate system, representing examples of clusters forming in column topology.
- FIG. 4 b shows examples of defect clusters in which the centroids of the density shape model, or centers of density mass, align in a y-axis in a Cartesian coordinate, representing examples of clusters forming in row topology.
- FIG. 4 c shows examples of defect clusters in which the centroids form a ring with centers of density mass equidistant from the center of the wafer and with the spines tangential to the ring, representing examples of annular defect clusters forming in annular topology.
- FIG. 4 a shows examples of defect clusters in which the centroids form a ring with centers of density mass equidistant from the center of the wafer and with the spines tangential to the ring, representing examples of annular defect clusters forming in annular top
- FIG. 4 d shows examples of defect clusters in which the centroids form along a radius of the wafer and the orientation of the spines are on the same axis as the radius, representing radial defect clusters forming in radial topology.
- FIG. 4 e shows examples of scratch defect clusters in which the spines of the density shape model are coincident within a specified tolerance with a line connecting the centroids of the same density shape models, representing examples of patterned scratch clusters forming in off-center linear topology.
- FIG. 4 f shows examples of defect clusters which spiral out from the center of wafer, which is based on the change of rotational speed as a function of sweep angle and the distance of the center of the mass to the center of the wafer, representing examples of spiral clusters forming in spiral topology.
- Step 206 allows the user to review the features generated by steps 204 and 205 to and apply a tolerance value for one or more features.
- the user may apply a 100 micron tolerance value against the trained model's length of the spine which is 1000 microns. That means that clusters with a spine greater than 900 microns and less than 1100 microns will satisfy the rule.
- Present embodiments perform pattern recognition using the broad class of if-then rules for classification of defect clusters defined in steps 207 and 208 and stored in a library 300 .
- a very simple example of an if-then rule is as follows: IF (SpineLength( x )>2000 microns) AND (Width( x ) ⁇ 10 microns) THEN ThinScratch( x ) This means that if the density shape model defined by a defect cluster x has a spine length greater than 2000 microns and a width less than 10 microns, then it is declared to be a thin scratch.
- a predicate such as IsObjectOrthogonal, comprises a test that returns a value of logical True or False.
- a function, such as SpineLength(x) comprises a test that returns a numerical value as illustrated above. If-then rules can incorporate predicates as well as functions. Step 200 describes learning such if-then rules.
- FIG. 3 is a flow diagram illustrating a method for zonal analysis and rule extraction for step 250 .
- the rule extraction process 250 is described in terms of a graphical user interface, operating within a graphical computer operating system.
- Step 252 displays a wafer defect map with ring or annular type defects used for learning such if-then rules.
- Step 253 allows a user, using a pointing device or equivalent device, to draw the boundary of a zone used for training such if-then rules.
- FIG. 3 a shows a graphical user interface of a zone editor 90 with annular defect cluster of rings 97 superimposed on the defects, according to an embodiment of the present invention.
- the zone editor 90 comprises one or more of the following tools to define or draw a zone over the wafer defect map:
- the zonal analysis focuses on pattern recognition of zonal defect clusters using if-then rules for classification of a defect zone or multiple zones, defined in steps 254 , 257 , 258 and stored in a library 300 .
- a very simple if-then rule example to define an annular defect cluster comprises: IF (PercentageOfZoneDefectsOverTotalWaferDefects( x )>70) AND (PercentageOfDefectiveDiesOverTheTotalDiesInZone( x )>30) THEN TRUE
- Step 250 describes the learning of such if-then rules.
- Step 400 may be deployed in automated defect classification.
- the rules stored in the library 300 can be used to classify spatial defect clusters during the fabrication of wafers, by taking an inspection results file and applying the clustering and if-then rules to determine whether there are any defect combinations within the inspection results file that match the rules specified in the signature library.
- a feedback loop 500 is incorporated to allow the user to optimize the rule or to feedback defect clusters that have been misclassified by previous rules.
- DFM software One of the objectives of DFM software is to provide feedback from the manufacturing process to the design engineer. For example, a chip designer may be interested in analyzing “hot spots,” those areas of the device that contain a disproportionately high number of wafer inspection defects. Some of these hot spots may result from a chip design that is particularly difficult to manufacture. To obtain this type of information, a large volume of wafer inspection data needs to be processed.
- Existing DFM software typically create a die stack from large numbers of wafer inspection files, and use a thresholding function to determine the location of single points that have a high percentage of defects. These points may then be overlaid onto the design data. As the threshold is lowered, an increasing number of points exceed the threshold, making it more difficult to analyze.
- single point failures there may be other larger systematic defect distributions of interest that can best be identified using the above described techniques of spatial signature analysis (SSA).
- SSA spatial signature analysis
- the single point defect densities of the distribution may not exceed the threshold; however, the spatial attributes of the distribution may allow it to be identified.
- the SSA software overlays the dies or reticles from a large number of wafer inspection levels, and uses die or reticle object analysis to identify specific object signatures that are of interest to the design engineer. These objects can then be converted to the same format as the design data so that they can be directly overlaid on the design data. In this way the designer can more easily identify systematic distributions of interest.
- Identifying the die stack or reticle stack cluster follows the above described steps 203 , 204 and 205 .
- the user examines a die stack or reticle stack.
- the user views the die or reticle stack while trying one or more combinations of density and distance parameters in order to capture desired stack clusters. Rules are extracted in the same fashion as with a wafer map. Once defects in the die or reticle stack cluster are identified, they can be put back into their original locations on the wafer map or, in the case of DFM analysis, they can be incorporated into the design files.
Abstract
Description
wherein
-
- xp, yp are the coordinates of the points piεP in the cluster;
- ωx(pi), ωy(pi) are the weight functions by x and y coordinates of point Pi;
- E is the best-fit ellipse.
This best-fit ellipse E has the center at the mass-center (x0, y0) of the cluster and has the same central second moment as the cluster, as shown by the following equation:
-
- a)
Centroid - b)
Spine - c)
Cross section width - d) Cross section symmetry. This may be symmetric 57, 64, 67 or
asymmetric 72. - e) Degree curvature of spine. This may be straight 56, 68, 71 or curved 61.
- f) The change of cross section as a function of sweep. This may be constant 62, 67, expanding 72, contracting, expanding then contracting 57, or contracting then expanding.
- g) The convex set is defined as any nonempty defect cluster (S) in which a line segment joining any two points in S always lies entirely within S. The shaded sets in
FIGS. 2 g (101), (102), (103), (104), (105) and (106) are convex sets. The shaded sets inFIGS. 2 g (107), 108 and (109) are not convex sets, since a line can be drawn, joining the indicated points within the set, which does not lie entirely within the set.
- a)
-
- a) Location of the density shape model on the wafer defect map, for example expressed in angle degrees relative to the wafer notch.
- b) Location of the density shape model on the wafer defect map in micron units, relative to the position of the center of the wafer defect map or as a function of the radius of the wafer.
- c) Orientation of the density shape spine in angle degrees, relative to the axis of the wafer.
- d) Tangential angle of the density shape spine in degrees as a function of the radius of the wafer.
- e) Elongation of the density shape model as a function of the major and minor axes.
- f) Roundness of the density shape model as a function of the shape.
- g) Density and distribution uniformity of the defects within the cluster boundary.
- h) Texture of the defect distribution with the cluster boundary.
IF (SpineLength(x)>2000 microns) AND (Width(x)<10 microns) THEN ThinScratch(x)
This means that if the density shape model defined by a defect cluster x has a spine length greater than 2000 microns and a width less than 10 microns, then it is declared to be a thin scratch. A predicate, such as IsObjectOrthogonal, comprises a test that returns a value of logical True or False. A function, such as SpineLength(x), comprises a test that returns a numerical value as illustrated above. If-then rules can incorporate predicates as well as functions. Step 200 describes learning such if-then rules.
-
- a) Quadrants 91
- b) Pie sections 92
- c) Rings 93
- d) Rectangles 94
- e)
Ellipses 95 - f)
Polygons 96
IF (PercentageOfZoneDefectsOverTotalWaferDefects(x)>70) AND (PercentageOfDefectiveDiesOverTheTotalDiesInZone(x)>30) THEN TRUE
Claims (26)
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
US11/849,869 US7987150B1 (en) | 2007-02-09 | 2007-09-04 | Method and apparatus for automated rule-based sourcing of substrate microfabrication defects |
Applications Claiming Priority (2)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
US90055107P | 2007-02-09 | 2007-02-09 | |
US11/849,869 US7987150B1 (en) | 2007-02-09 | 2007-09-04 | Method and apparatus for automated rule-based sourcing of substrate microfabrication defects |
Publications (1)
Publication Number | Publication Date |
---|---|
US7987150B1 true US7987150B1 (en) | 2011-07-26 |
Family
ID=44280183
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
US11/849,869 Expired - Fee Related US7987150B1 (en) | 2007-02-09 | 2007-09-04 | Method and apparatus for automated rule-based sourcing of substrate microfabrication defects |
Country Status (1)
Country | Link |
---|---|
US (1) | US7987150B1 (en) |
Cited By (12)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20130170733A1 (en) * | 2011-12-28 | 2013-07-04 | Elitetech Technology Co.,Ltd. | Method for building rule of thumb of defect classification, and methods for classifying defect and judging killer defect based on rule of thumb and critical area analysis |
US20140358830A1 (en) * | 2013-05-30 | 2014-12-04 | Synopsys, Inc. | Lithographic hotspot detection using multiple machine learning kernels |
CN107145595A (en) * | 2017-05-27 | 2017-09-08 | 南京英斯特网络科技有限公司 | A kind of mass data aggregation display method based on map |
US20190250060A1 (en) * | 2018-02-09 | 2019-08-15 | Olympus Corporation | Pipeline inspection information managing device and pipeline inspection information managing method |
WO2019180714A1 (en) * | 2018-03-22 | 2019-09-26 | Optimal Plus Ltd. | Methods and systems for detecting defects on an electronic assembly |
CN111448453A (en) * | 2017-10-06 | 2020-07-24 | 罗森瑞士股份公司 | Method for determining the geometry of a defect and method for determining the load capacity limit |
US20210142122A1 (en) * | 2019-10-06 | 2021-05-13 | Pdf Solutions, Inc. | Collaborative Learning Model for Semiconductor Applications |
US11024022B2 (en) * | 2017-03-03 | 2021-06-01 | Fujitsu Limited | Data generation method and data generation device |
CN113302478A (en) * | 2019-01-10 | 2021-08-24 | 朗姆研究公司 | Defect classification and source analysis for semiconductor devices |
US11120302B2 (en) | 2017-03-03 | 2021-09-14 | Fujitsu Limited | Data generation apparatus, data generation method, and non-transitory computer-readable storage medium for storing program |
US11158042B2 (en) * | 2019-07-10 | 2021-10-26 | International Business Machines Corporation | Object defect detection |
US11238189B2 (en) * | 2014-02-12 | 2022-02-01 | Asml Netherlands B.V. | Process window optimizer |
Citations (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20070288185A1 (en) * | 2003-12-31 | 2007-12-13 | Richard Burch | Method and System for Failure Signal Detention Analysis |
-
2007
- 2007-09-04 US US11/849,869 patent/US7987150B1/en not_active Expired - Fee Related
Patent Citations (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20070288185A1 (en) * | 2003-12-31 | 2007-12-13 | Richard Burch | Method and System for Failure Signal Detention Analysis |
Non-Patent Citations (4)
Title |
---|
Hugo A. D. do Nascimento et al. "Interactive graph clustering based on user hints", 2001, Lecturer of the Instituto de Informática, UFG-Brazil, 7 pages. * |
Lin "Parametric Wafer Map Visualization", IEEE Computer Graphics and Applications, 1999, pp. 1-4. * |
Tobin et al. "Integrated applications of inspection data in the semiconductor manufacturing environment," in Metrology-based Control for Micro-Manufacturing, vol. 4275 of SPIE Proceedings, 2001, pp. 31-40. * |
Tyagi et al. "Defect Clustering Viewed Through Generalized Poisson Distribution", IEEE, SM, vol. 5, No. 3, 1992, pp. 196-206. * |
Cited By (18)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US8908957B2 (en) * | 2011-12-28 | 2014-12-09 | Elitetech Technology Co.,Ltd. | Method for building rule of thumb of defect classification, and methods for classifying defect and judging killer defect based on rule of thumb and critical area analysis |
US20130170733A1 (en) * | 2011-12-28 | 2013-07-04 | Elitetech Technology Co.,Ltd. | Method for building rule of thumb of defect classification, and methods for classifying defect and judging killer defect based on rule of thumb and critical area analysis |
US20140358830A1 (en) * | 2013-05-30 | 2014-12-04 | Synopsys, Inc. | Lithographic hotspot detection using multiple machine learning kernels |
US11403564B2 (en) | 2013-05-30 | 2022-08-02 | Synopsys, Inc. | Lithographic hotspot detection using multiple machine learning kernels |
US20220147665A1 (en) * | 2014-02-12 | 2022-05-12 | Asml Netherlands B.V. | Process window optimizer |
US11238189B2 (en) * | 2014-02-12 | 2022-02-01 | Asml Netherlands B.V. | Process window optimizer |
US11120302B2 (en) | 2017-03-03 | 2021-09-14 | Fujitsu Limited | Data generation apparatus, data generation method, and non-transitory computer-readable storage medium for storing program |
US11024022B2 (en) * | 2017-03-03 | 2021-06-01 | Fujitsu Limited | Data generation method and data generation device |
CN107145595A (en) * | 2017-05-27 | 2017-09-08 | 南京英斯特网络科技有限公司 | A kind of mass data aggregation display method based on map |
CN111448453A (en) * | 2017-10-06 | 2020-07-24 | 罗森瑞士股份公司 | Method for determining the geometry of a defect and method for determining the load capacity limit |
US20190250060A1 (en) * | 2018-02-09 | 2019-08-15 | Olympus Corporation | Pipeline inspection information managing device and pipeline inspection information managing method |
US11852684B2 (en) * | 2018-03-22 | 2023-12-26 | Optimal Plus Ltd. | Methods and systems for detecting defects on an electronic assembly |
US20210041501A1 (en) * | 2018-03-22 | 2021-02-11 | Optimal Plus Ltd. | Methods and systems for detecting defects on an electronic assembly |
WO2019180714A1 (en) * | 2018-03-22 | 2019-09-26 | Optimal Plus Ltd. | Methods and systems for detecting defects on an electronic assembly |
US11650250B2 (en) * | 2018-03-22 | 2023-05-16 | Optimal Plus Ltd. | Methods and systems for detecting defects on an electronic assembly |
CN113302478A (en) * | 2019-01-10 | 2021-08-24 | 朗姆研究公司 | Defect classification and source analysis for semiconductor devices |
US11158042B2 (en) * | 2019-07-10 | 2021-10-26 | International Business Machines Corporation | Object defect detection |
US20210142122A1 (en) * | 2019-10-06 | 2021-05-13 | Pdf Solutions, Inc. | Collaborative Learning Model for Semiconductor Applications |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
US7987150B1 (en) | Method and apparatus for automated rule-based sourcing of substrate microfabrication defects | |
US7602962B2 (en) | Method of classifying defects using multiple inspection machines | |
Fan et al. | Wafer defect patterns recognition based on OPTICS and multi-label classification | |
US8331651B2 (en) | Method and apparatus for inspecting defect of pattern formed on semiconductor device | |
JP5429869B2 (en) | Pattern inspection apparatus and method | |
US7231079B2 (en) | Method and system for inspecting electronic circuit pattern | |
US6775817B2 (en) | Inspection system and semiconductor device manufacturing method | |
US9165843B2 (en) | Systems and methods of automatically detecting failure patterns for semiconductor wafer fabrication processes | |
JP5599387B2 (en) | System and method for detecting defects on a wafer and generating inspection results | |
TWI755613B (en) | Pattern grouping method based on machine learning | |
JP4787673B2 (en) | Pattern inspection apparatus and method | |
WO2004111618A2 (en) | Method and system for classifiying defects occurring at a surface of a substrate using graphical representation of multi-channel data | |
JP2002100660A (en) | Defect detecting method, defect observing method and defect detecting apparatus | |
US6687633B2 (en) | Inspection system, inspection apparatus, inspection program, and production method of semiconductor devices | |
TWI773063B (en) | Image generation system and image generation method | |
CN113439276A (en) | Defect classification in semiconductor samples based on machine learning | |
US7229845B1 (en) | Automated sourcing of substrate microfabrication defects using defects signatures | |
EP4128141A1 (en) | Examination of a semiconductor specimen | |
EP4174766A1 (en) | Generating segmentation masks for training a semantic segmentation model | |
JP4597509B2 (en) | Pattern inspection apparatus and pattern inspection method | |
CN116935092A (en) | Automated defect classification and detection | |
US20090080763A1 (en) | Method and system for the visual classification of defects | |
US20220334567A1 (en) | Fabrication fingerprint for proactive yield management | |
CN113763312B (en) | Detection of defects in semiconductor samples using weak labels | |
TW202338894A (en) | Measurement method and apparatus for semiconductor features with increased throughput |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
AS | Assignment |
Owner name: SIGLAZ, CALIFORNIA Free format text: ASSIGNMENT OF ASSIGNORS INTEREST;ASSIGNORS:LUU, VICTOR V;POREDA, JOHN;TRINH, THIEU NGUYEN;AND OTHERS;SIGNING DATES FROM 20070827 TO 20070829;REEL/FRAME:019778/0872 |
|
STCF | Information on status: patent grant |
Free format text: PATENTED CASE |
|
FPAY | Fee payment |
Year of fee payment: 4 |
|
FEPP | Fee payment procedure |
Free format text: MAINTENANCE FEE REMINDER MAILED (ORIGINAL EVENT CODE: REM.); ENTITY STATUS OF PATENT OWNER: SMALL ENTITY |
|
LAPS | Lapse for failure to pay maintenance fees |
Free format text: PATENT EXPIRED FOR FAILURE TO PAY MAINTENANCE FEES (ORIGINAL EVENT CODE: EXP.); ENTITY STATUS OF PATENT OWNER: SMALL ENTITY |
|
STCH | Information on status: patent discontinuation |
Free format text: PATENT EXPIRED DUE TO NONPAYMENT OF MAINTENANCE FEES UNDER 37 CFR 1.362 |
|
FP | Lapsed due to failure to pay maintenance fee |
Effective date: 20190726 |